The past 25 yr have witnessed a sea change in attitudes and legislation regarding driving under the influence (DUI) of drugs and alcohol. Although the first laws against drunk driving were passed as long ago as 1910 by New York and by California in 1911, statutes in most states prior to about 1980 simply prohibited "driving while intoxicated," without providing specific guidelines to define "intoxicated." As a result, much discretion was involved in the arrest and prosecution of drinking drivers, and punishment was often light, even for chronic offenders.
In the early 1980s, however, under pressure from insurance companies and advocacy groups such as Mothers Against Drunk Driving (MADD) and accompanied by threats of loss of federal highway funds for noncompliance, all states passed tougher drunk driving statutes, often incorporating explicit measures of presumed alcohol impairment. In 1980, for example, only 15 states had legislation establishing a permissible blood alcohol content (BAC), with 0.10 g/dL then defined as the presumed level of intoxication. By 2005, after a renewed push again involving pressure from highway funding, all states had a BAC limit at a lower threshold of 0.08 g/dL.
BAC laws have been accompanied by a raft of other alcohol-control measures: MADD (2005) lists 39 pieces of recommended state legislation for the prevention of drunk driving, ranging from administrative license revocation (ALR) (1) to zero tolerance (ZT) of drinking by minor drivers; by 2005, the median state had passed 27 pieces, and all had passed at least 17.
Clearly, progress has been made in getting drunk drivers off the streets. Advances in automotive safety engineering over time have led to declines in fatalities from all causes, but from 1982, the first year for which reliable measures of victims' BAC have been collected, to 2004, the National Highway Traffic Safety Administration (NHTSA, 2005a) reports that the number of alcohol-related traffic fatalities has fallen from 26, 173 to 16,694, even as the number of miles traveled has increased by 81%. More important for measuring the effect of control legislation, the rate of alcohol involvement in fatal crashes fell from 60% in 1982 to 39% in 2004.
The good news over the past 25 yr is tempered, however, by slow progress in the past decade in reducing total fatalities and almost no progress in further reducing the rate of alcohol involvement. Figure 1 displays the trends in total fatalities per 100,000 population (solid line, left axis) for the years 1975--2004 and the percentage of fatalities that are alcohol related (dotted line, right axis) for the years 1982-2004. As shown in the chart, total fatality rates were constant until 1980, fell sharply during the 1980-1982 recession, stabilized, and then fell sharply again during the period 1987-1992. After 1992, fatality rates continued to decline but at a much slower pace than in the previous decade, with the number of total traffic deaths fluctuating around 40,000 annually for the past decade.
[FIGURE 1 OMITTED]
Alcohol-related fatalities as a share of the total have only declined by 1% since 1997. Other measures of alcohol-impaired driving, including surveys of self-reported drinking and driving by the Centers for Disease Control, cited in Quinlan et al. (2005), and by NHTSA (2003a), report actual increases from 1997 to 2002 in the frequency of DUI and in the number of drinks consumed when driving after drinking. These developments lead to questions regarding the effectiveness of the many types of drunk driving laws passed in recent years, including the push to reduce the BAC limit to 0.08.
Figure 2 displays the 5-yr frequencies of the adoption by the states since 1980 of two important types of alcohol control legislation, BAC laws and ALR laws. The bottom of each bar (shaded black) is the number of states adopting ALR laws during the 5-yr period, the middle (in gray) is the number adopting BAC 10 laws, and the top (in white) is the number adopting BAC 08 laws. …